Persistent Disks Cheat Sheet

• Are durable network storage devices that you can provision to host your virtual machine instances.

Persistent Disks Features

• Data on each persistent disk is distributed across several physical disks and is designed for high durability. It stores data redundantly to ensure data integrity.
• Persistent disks are resizable to accommodate larger storage requirements.
• It can be attached to virtual machines running on Compute Engine (GCE) or Google Kubernetes Engine (GKE).

• You cannot attach a persistent disk to an instance on another project.
• Your storage is independent of your virtual machine instances so you can detach or move your PDs to keep your data even after you delete your instances.
• You can only change the size of a Persistent Disk incrementally.
• Hyperdisk: Google’s next-generation block storage. Offers higher performance limits and lets you independently configure IOPS, throughput, and capacity (unlike Persistent Disk where performance scales with size).
  • Hyperdisk Balanced: For general-purpose workloads.
  • Hyperdisk Extreme: For high-performance databases and mission-critical apps.
  • Hyperdisk Throughput: For large-scale, throughput-intensive workloads.
  • Hyperdisk ML: Optimized for AI/ML model training and inference.
  • Hyperdisk Balanced High Availability: Regionally replicated for critical enterprise apps.
• Asynchronous Replication: Replicate disk data to a secondary region for disaster recovery with low Recovery Point Objective (RPO) and Recovery Time Objective (RTO).

Zonal and Regional Persistent Disks

You can configure your PD to be zonal or regional.

• Zonal Disks
  • Are relatively faster than Regional disks and are found in a zone.
• Regional Disks
  • Provides replication of data between two zones in the same region.
  • Is designed for building robust and highly available systems on Compute Engine.

Persistent Disk & Hyperdisk Types

• Standard (pd-standard):
  • Backed by HDD. Efficient, economical for sequential read/write. Not optimized for high random IOPS.
• Balanced (pd-balanced):
  • Backed by SSD. Balance of performance and cost for general-purpose workloads.
• SSD (pd-ssd):
  • Backed by SSD. Designed for single-digit millisecond latencies and high random IOPS.
• Extreme (pd-extreme):
  • Backed by SSD. Highest performance Persistent Disk for mission-critical databases. You provision specific IOPS.
• Hyperdisk Balanced:
  • Independently configure IOPS, throughput, and capacity. For general-purpose workloads.
• Hyperdisk Extreme:
  • Independently configure IOPS and capacity. Highest performance for mission-critical databases.
• Hyperdisk Throughput:
  • Independently configure throughput and capacity. For large-scale, throughput-intensive workloads.
• Hyperdisk ML:
  • Optimized for AI/ML workloads with configurable throughput.
• Hyperdisk Balanced High Availability:
  • Regionally replicated version of Hyperdisk Balanced for enterprise apps requiring high availability.

Encryption

• Data on persistent disks are automatically encrypted at rest and in transit by system defined encryption keys or with customer-supplied keys.
• To control your data encryption, you can create PDs with your own encryption keys.

Snapshots

• Standard snapshots: Incremental backups stored remotely. Use for long-term backup and disaster recovery. Billed for storage used.
• Archive snapshots: More cost-efficient for long-term data retention. Billed for storage (minimum 90 days) plus retrieval fees when restoring.
• Instant snapshots: Near-instantaneous, point-in-time backups stored locally (in same zone). Ideal for short-term protection against data corruption or user error. Charged per operation and for differential data stored.
• Snapshot schedules: Automate regular backups of your disks.
• Snapshots are global resources by default; can also be created as regionally scoped (Preview).

Pricing

Provisioning persistent disks incurs cost based on the following factors:

• Disk Storage:
  • Standard: Billed per GiB per month (first 30 GiB free)
  • Balanced: Billed per GiB per month
  • SSD: Billed per GiB per month
  • Extreme: Billed per GiB per month plus provisioned IOPS charges
  • Regional disks: Approximately 2x cost of zonal (replicated across two zones)
  • Hyperdisk: Includes provisioned capacity plus separate charges for provisioned IOPS/throughput (varies by type)
• Snapshot Storage:
  • Standard snapshots: Billed per GiB per month (incremental, actual data stored)
  • Archive snapshots: Billed per GiB per month (minimum 90-day storage) plus retrieval fees when restoring
  • Instant snapshots: Per-operation charge plus storage for differential data (at same rate as source disk type)
• Network Charges:
  • Creating snapshot from disk to same region: No charge
  • Creating snapshot to different region or multi-region: Inter-region egress rates apply
  • Restoring snapshot to disk in different region: Inter-region egress rates apply
• Asynchronous Replication:
  • Replication Protection: Charged per GiB-month for primary and secondary disks (based on provisioned capacity)
  • Network transfer: Charged per GiB of data replicated between regions
• Image Storage:
  • Custom images: Billed per GiB per month
  • Machine images: Billed per GiB per month
• Premium Images (additional hourly license costs beyond base VM charges):
  • RHEL and RHEL for SAP: Per-core per hour (core-hour model effective July 1, 2024)
  • SLES and SLES for SAP: Per hour (1-minute minimum, per-second increments after)
  • Ubuntu Pro: Per GB RAM per hour plus vCPU tiered pricing
  • Windows Server: Per hour per vCPU
  • SQL Server: Per core per hour (4-core minimum)

All disk-related charges are prorated per second. Spot VMs also receive discounted rates for attached disks.

For current pricing details, refer to the official Google Cloud Compute Engine disk and image pricing page.

Persistent Disks Cheat Sheet References:

https://cloud.google.com/compute/docs/disks
https://cloud.google.com/compute/docs/disks/create-snapshots
https://cloud.google.com/compute/disks-image-pricing